On Sep 25, 2013, at 1:00 PM, Mouse <mouse at rodents-montreal.org> wrote:
+-----------+-------- +5V
| |
4.7K Ohm Collector
| .
D33D1 --- 220ohm ------+-------- Base Q1, PNP Silicon
.
Emitter ----- 47 Ohm ---+--- RS-232 OUT
|
3.3 K Ohm
|
-12V (Regulated)
Sure you don't have collector and emitter switched on that diagram? If
you don't, you're using it "backwards"; lots of BJTs will work if you
swap C and E, but not as well as they will the right way around. (The
usual simplistic diagrams of BJTs make them look symmetric in this
regard, but as I understand it the diagrams are a simplification and
the actual geometry is not as symmetric as the diagrams make it look.)
The asymmetry is, in fact, the only thing that makes one end a
collector and the other end an emitter. If it were truly
symmetrical, it genuinely wouldn't matter which way you placed it.
But yes, it will (sorta) work placed backwards; I've had a number
of maddening debugging sessions that all boiled down to "I forgot
the standard pin ordering on a TO92 package", since the transistor
will *work* backwards, but is much weaker than expected.
- Dave